South Africa is home to the highest number of people living with HIV/AIDS in the world with an adult (15-49 years) population prevalence of 17.8%. The South African HIV epidemic is one of the most disparate health conditions, with young black men and women disproportionately affected. The historical and structural factors underpinning these disparities continue to shape the structure of social and sexual networks which have varying consequences for HIV risk and transmission. Despite its central role in HIV risk disparities, work around understanding social and sexual network attributes that place certain individuals at a heightened risk of infection in South Africa is limited. We addressed these gaps by exploring three key aspects of network structure among men who have sex with men (MSM) and young black women in South Africa.
In the first study we created five social network typologies of MSM based on varying levels of connectivity between friends, family and sex partners. These typologies were characterized in relation to sexual risk-taking and provided insight into how differences in social network structure influence the way in which sexual risk manifests among this key population.
The second study compared race-assortative sexual mixing patterns of MSM using Newman's assortativity coefficient and the E-I statistic. Race assortativity was significantly higher than would have been expected under conditions of random mixing, while comparison of race-specific E-I statistics revealed heterogeneity in race assortativity by ego race.
The final study explored the context of age-disparate sexual partnerships among young women (15-29 years) in rural Kwazulu-Natal by evaluating both individual- and community-level factors within a multilevel modeling framework. The community ratio of unmarried females to unmarried males, female age and marital status were significant predictors of engaging in an age-disparate partnership.
Taken together, these studies demonstrate the utility of evaluating social and sexual network structure in understanding the existence of HIV risk disparities.
Table of Contents
Chapter 1: Background and Significance ..1
Striking disparities in HIV risk exist in the South African HIV epidemic ..1
HIV risk among men who have sex with men in South Africa ..2
Disparities in HIV risk among women in South Africa 4
Network-based approaches to better understand HIV risk disparities 5
Network studies in South Africa 8
Chapter 2: Social network typologies and sexual risk-taking among men who have sex with men in Cape Town and Port Elizabeth, South Africa ..11
Study Population and Recruitment ..16
Study Procedures ..17
Data Analysis ..17
Chapter 3: Race assortativity among men who have sex with men in Cape Town, South Africa ..32
Chapter 4: Understanding the context of age-disparate relationships among young women in rural South Africa ..52
Community-level determinants of age-disparate sexual relationships ..56
Chapter 5: Conclusions and Future Directions ..75
A1. Code for running a permutation test by constructing a study population under the null hypothesis of random mixing of partners (SAS) ..98
List of Tables
Table 1: Description of the Social Network Typologies of men who have sex with men (MSM) in Cape Town and Port Elizabeth, as determined by the level of connectivity between social network subgroups i.e., friends, family members and sex partners ..30
Table 2: Demographic and behavioral characteristics of 78 men who have sex with men (MSM) participants in Cape Town and Port Elizabeth, South Africa ..31
Table 3: Individual-level demographic and behavioral characteristics of 314 MSM participants in Cape Town, South Africa ..46
Table 4: Characteristics of 714 partnerships reported by 201 black, 82 coloured and 15 white MSM in Cape Town, South Africa ..47
Table 5: Ego and dyad-level demographic and behavioral characteristics by race assortativity ..48
Table 6: Contingency tables by ego and alter race stratified by demographic and behavioral characteristics ..49
Table 7: Community-level variables and corresponding domains included in a multilevel logistic regression model evaluating the context of women (15-29 years) engaging in an age-disparate relationship ..68
Table 8: Baseline characteristics of study sample of 15- to 29-year old women (n=8290) ..69
Table 9: Selected community-level characteristics by neighborhood type ..71
Table 10: Unconditional multilevel logistic regression model for assessing associations among women engaging in age-disparate relationships ..72
Table 11: Odds ratios (and 95% confidence intervals) from multilevel logistic regression model assessing associations among women engaging in age-disparate relationships ..73
List of Figures
Figure 1: Mixing matrix cross-tabulating ego race with alter race ..38
Figure 2: Age disparity between female respondent (15-29 years old) and her most recent male sex partner ..67
About this Dissertation
|Committee Chair / Thesis Advisor|
|Using social and sexual network structure to model HIV risk disparities in South Africa ()||2018-08-28||